Forensic Engineering Underground

ROCK MECHANICS IN UNDERGROUND CONSTRUCTION (With CD-ROM)- Proceedings of the ISRM International Symposium 2006 and the 4th Asian Rock Mechanics Symposium© World Scientific Publishing Co. Pte. Ltd.http://www.worldscibooks.com/engineering/6335.html

FORENSIC ENGINEERING FOR UNDERGROUND CONSTRUCTION

E. T. BROWN

Golder Associates Pty. Ltd., Brisbane, Australia

([email protected])

In the context of underground construction, forensic engineering is taken to be the application of engineering

principles and methodologies to determine the cause of a performance deficiency, often a collapse, in an

excavation, and the reporting of the findings, usually in the form of an expert opinion within the legal system.

The procedures that may be used in forensic geotechnical investigations and the interface of the engineer with

the legal system are discussed. The application of the principles and methodologies outlined are illustrated

through a brief account of the investigation of the collapse of a small part of an excavation in the Lane Cove

Tunnel Project, Sydney, Australia, on 2 November, 2005.

Keywords: Accident investigation; failure; forensic engineering; mining; risk; tunnelling; underground

construction.

1. Introduction

In common with other areas of geotechnical engineering practice, underground construction

involves a number of uncertainties and risks, many of them associated with the inherent variability

and unknown properties of the geological materials involved. These and other factors may lead to

deficiencies in excavation performance, to the collapse of an excavation and, on occasion, to the

loss of life.

It is inevitable that, when an underground failure or collapse has occurred, either in civil

engineering construction or in mining, an investigation will be carried out into the causes of the

failure. Depending on the nature and severity of the failure, this investigation will be carried out, at

least in part, by suitably experienced specialist engineers. In many cases, some form of legal

proceedings will follow, either to determine the causes of damage, loss of income or, in some

cases, the loss of life, and/or to resolve contractual and responsibility issues and allocate costs.

This process will also involve specialist engineers, usually as expert witnesses. The professional

engineering work carried out in these cases has come to be described as forensic engineering.

This paper discusses the general nature of forensic engineering and the special issues and

difficulties confronting forensic geotechnical engineers, particularly in the investigation of failures

or collapses in underground construction and underground mining. The investigation methods used

are outlined and the important interface with legal systems is discussed. In the author’s experience

of forensic investigations in underground civil construction and mining, this legal aspect of the

forensic engineer’s role is becoming increasingly important and demanding, given the increasing

proclivity of some authorities to prosecute engineers and their employers in the courts. Finally, a

brief account is given of the author’s investigation of the collapse of a small section of an

excavation in the Lane Cove Tunnel Project in Sydney, New South Wales, Australia, on 2

November 2005.

2. The General Nature of Forensic Engineering

Technological innovation and advances in engineering have always been attended by failure of

one type or another, including the quite spectacular collapse of structures such as bridges and


dams (e.g. Lewis, 2004; Petroski, 1985, 1994). More recently, as financial losses and the loss of

reputation have increased, personal injury and the loss of life have come to be regraded more

seriously than they sometimes were in the past, and society has become generally more litigious, an

area of engineering practice known as forensic engineering has developed. Forensic engineering

now has its own specialist societies, consulting firms, conferences, literature and university

courses, and has attracted popular attention through television programs and books (e.g. Lewis,

2004; Wearne, 1999).

A number of definitions of forensic engineering are available in the literature. For example,

Specter (1987) defines forensic engineering as “the art and science of professional practice of

those qualified to serve as engineering experts in matters before courts of law or in arbitration

proceedings”. Similarly, Noon (2001) defines forensic engineering as “the application of

engineering principles, knowledge, skills, and methodologies to answer questions of fact that may

have legal ramifications”. Carper (2000) says that “the forensic engineer is a professional

engineer who deals with the engineering aspects of legal problems. Activities associated with

forensic engineering include determination of the physical or technical causes of accidents or

failures, preparation of reports, and presentation of testimony or advisory opinions that assist in

resolution of related disputes. The forensic engineer may also be asked to render an opinion

regarding responsibility for the accident or failure”.

Following Lewis (2003) and Noon (2001), in the context of underground construction,

forensic engineering will be taken here to be the application of engineering principles and

methodologies to determine the cause of a performance deficiency, usually a collapse, in an

excavation, and the reporting of the findings, usually in the form of an expert opinion within the

legal system. Some uses of the term forensic engineering do not reflect this involvement with the

legal system. However, this legal element is central to the definition, recognition and practice of

forensic engineering in some countries (e.g. Australia, USA), and will form an essential part of the

discussion presented here.

Forensic engineering as defined above is concerned typically with investigations of failures of

constructed facilities; rock falls, excavation collapses and other accidents in mines; fires and

explosions; air and rail crashes; aspects of traffic accidents; and failures of consumer products.

The more serious events of these types can lead to significant injury and to fatalities as well as to

financial loss.

Forensic engineering investigations involve a number of steps. In general, the forensic

engineer collects evidence of several types and then carries out analyses, again of various types, to

determine the “who, what, where, when, why and how” of the deficient performance or failure of

engineered facilities, systems and products, including accidents. A range of formal and less formal

procedures may be used to guide the investigations (e.g. Greenspan et al., 1989; Lewis, 2003;

Noon, 2001). Figure 1 shows the steps typically used in forensic engineering investigations within

a civil engineering context. The legal terminology used is that applying in the USA.

Communicating the results is a vitally important stage of the investigation. This

communication may be required, not only to facility owners, contractors and other professional

engineers, but also in reports to lawyers and statutory bodies, in expert witness statements in legal

proceedings, and in statements to the press and the public. In many cases, the expert’s report may

be confidential or protected by legal professional privilege. It is probable that only a low

percentage of cases for which forensic engineering investigations are undertaken and expert

witness reports are prepared, actually reach the courts (e.g. Brookes, 2006). However, the

information contained in the reports may be protected by legal privilege and remain confidential so


that the results cannot be published and the potentially valuable information that the reports

contain never reach the engineering profession.

Fig. 1. Flow chart of a typical forensic engineering investigation (after Greenspan et al., 1989).

The important question of what qualifies an engineer to be recognised as an “expert” is

discussed in some detail by a number of authorities including Carper (2001), Greenspan et al.

(1989) and Lewis (2003) who concludes that the key attributes of an expert engineer are

“education, training, experience, skill and knowledge” and that the engineer must be able to

“perform his or her work accurately, objectively and in a professional manner”. Forensic

engineers or engineers serving as expert witnesses have to be especially aware of the ethical

practice issues to be discussed in Section 5 below.

Client

Failure

Attorney

Forensic engineer Engineering agreement

Client interview Engineering agreement

Review of claims

Planning the investigation

Site observations

Interviews Sketches

Evidence: test samples measurements field data

Video Photos

Document search Research Team formation Experts

Testing laboratories

Analysis Conclusions

Report

Litigation

Expert testimony Litigation

Depositions Interrogatories


3. Forensic Geotechnical Engineering

3.1. Failures in Geotechnical Engineering

Engineering in the natural materials found on and under the surface of the Earth has long been

fraught with difficulty. Perhaps even more so than in other areas of engineering, there have always

been performance deficiencies and failures of geotechnical engineering projects. The most extreme

cases, generally arising from dam failures or major landslides, can cause significant damage to

property and infrastructure, and the loss of life (e.g. Wearne, 1999). Particular difficulties arise

from the inherent spatial variability of soil deposits and rock masses. Errors or omissions having

significant engineering consequences can be made in geological interpretations, there can be

variations in geotechnical properties over a project site, and apparently minor geological features

can have major influences on the performance of engineered structures (e.g. Terzaghi, 1929).

Attempts have been made over the last two or three decades to account for some of these factors

through probability-based reliability and risk analyses which have now become part of the corpus

of geotechnical engineering (e.g. Bea 2006; Christian, 2004; Einstein, 1996; Eskesen et al., 2004).

The concept of reliability as “the likelihood that a system will perform in an acceptable manner”

(Bea, 2006) is important in forensic geotechnical engineering.

It is not surprising, therefore, that studies and analyses of the failures of foundations, slopes,

dams and tunnels, for example, have been important in the development of geotechnical

engineering research, knowledge and practice. Following Leonards (1982), failure will be taken

here to be an “unacceptable difference between expected and observed performance”. Many

failures involve sometimes catastrophic instability which arises when some form of sudden rupture

develops. The geotechnical engineering literature is replete with detailed examples and broader

studies of the causes of geotechnical failures (e.g. Bea, 2006; Day, 1998; Leonards, 1982; Londe,

1987; Müller, 1968; Osterberg, 1989; Sowers, 1993). Several of these studies include

considerations of the influence of human factors on geotechnical engineering failures. For

example, Sowers’ (1993) study of more than 500 well-documented foundation failures showed that

the majority of the failures were due to “human shortcomings”. Only 12% of the failures studied

were attributed to the absence of relevant technical knowledge or solutions.

Generally, forensic geotechnical investigations follow the broad pattern illustrated by Figure 1

(Day, 1998). In many cases, careful and detailed re-investigation of the site is required with

detailed geological mapping, drilling, sampling and testing (e.g. Alonso and Gens, 2006a;

Skempton and Vaughan, 1993). In order to resolve some cases, similarly careful and detailed

analyses of the data, and back-analyses of the problem, are required (e.g. Alonso and Gens, 2006b;

Gens and Alonso, 2006; Potts et al., 1990). Although not always falling within the purview of

forensic engineering as defined here, similar forensic investigations may also be associated with

the repair and/or restitution of historic structures (as in the famous case of the Leaning Tower of

Pisa), with geoenvironmental problems, or in the aftermath of natural disasters such as

earthquakes. If the results of such investigations can be brought to the attention of the profession

through conference presentations or publication, they can make significant contributions to the

advancement of geotechnical engineering knowledge (e.g. Londe, 1987).

3.2. Underground Construction

Underground construction in soils and rocks can suffer from the same types of errors and

uncertainties as those outlined for geotechnical engineering more generally. Although fatalities can


occur and severe financial losses may result, the failures or collapses occurring in underground

construction do not often have the spectacular or disastrous effects of the major dam failures or

landslides referred to above. Tunnelling or the excavation of caverns may be slowed or halted by

frequent small or large falls of ground (e.g. Feld and Carper, 1997), squeezing ground conditions

(e.g. Hoek, 2001), groundwater problems (Hoek, 2001) including inundation, the sudden

development of very large settlements or sinkholes above tunnels or other underground

excavations (Shirlaw and Boone, 2005), or the unanticipated and continuing development of

excavation deformations (e.g. Stabel and Samani, 2003). Occurrences of the penultimate type have

occurred in a large number of projects, including railway tunnels in Singapore (Shirlaw et al.,

2003). Such occurrences can have damaging effects on buildings and on surface and near-surface

infrastructure, as in the case to be discussed in Section 6 below.

Underground mining in hard or soft rock can suffer from similar problems to those outlined

for underground construction. Although the purposes of many of the excavations made and of the

operations carried out in underground mining, specifically those associated with the stopes from

which the ore is extracted, differ from those in underground construction, modern large-scale

underground mining does have associated with it, large numbers of underground infrastructure and

transportation excavations which have elements in common with civil engineering excavations.

Unfortunately, fatalities arising from broadly geotechnical causes have been all too common in the

international mining industry, bringing with them legal proceedings of one type or another and the

need for forensic engineering investigations of the general type being discussed here. Rock

bursting, which is not unknown in civil construction, has been a particular cause of damage and

fatalities in deep, hard rock mining, most notably in the deep level gold mines of South Africa. In

addition to collapses underground, underground mining can cause subsidence and disruption to the

surface, damaging buildings and infrastructure. As in underground construction, throughout mining

history, there have been several major cases of the inundation of mine workings by water or

tailings leading to loss of life and of production.

The forensic investigations carried out in these various cases use the general principles and

methods discussed elsewhere in this paper. They also require knowledge of the principles of rock

mechanics and of the engineering behaviour of rocks and rock masses. De Ambrosis and Kotze

(2004) provide an excellent example of the detailed investigation of two large stress-induced roof

collapses that occurred in the roof strata of an underground LPG storage facility in Sydney, New

South Wales, Australia. The author has had experience of investigations of several of the types of

failure identified in this sub-section.

4. Formal Analysis Tools and Investigation Methods

4.1. Overview

A wide range of formal investigation methods and analysis tools and approaches are available for

use in forensic engineering investigations. Many of them have their basis in hazard identification,

risk assessment and risk management or control. Indeed, it can be argued that in underground

construction and rock engineering more generally, it is now possible to identify all of the

geotechnical hazards likely to have an impact on a project. The forensic engineering task then

becomes one of determining which of those hazards contributed to the incident being investigated

(Hudson, 2006).


The analytical tools available include a range of specific techniques such as causal analysis,

energy/barrier analysis, event trees, fault trees, human error analysis, Petri nets and sequentially

timed events plotting (STEP) (e.g. Hadipriono, 2002; Joy, 2004; Kontogiannis et al., 2000). These

specific techniques may be incorporated into more integrated analysis methods such as the Incident

Cause Analysis Method (ICAM) (Gibb et al., 2004) and overall investigation methods such as the

System Safety Accident Investigation (SSAI) approach (Joy, 2004).

These tools and approaches have been developed and applied to accidents and failures of a

range of types in a number of industries, including the construction and mining industries. The

extent to which they are applied in a given case will depend on the severity or consequences of the

incident concerned. Quite often, these tools and approaches may form the basis of the in-house

policies and protocols of companies or organizations. In other cases, particularly when major

catastrophic events occur, they may be used in accident investigations carried out by external

parties. The following outlines of ICAM and SSAI are presented in generic terms rather than in

terms that are specific to underground construction.

4.2. The Incident Cause Analysis Method (ICAM)

The Incident Cause Analysis Method (ICAM) is an analysis tool that sorts the findings of an

investigation into a structured framework. Its fundamental concept is the acceptance of the

inevitability of human error. It is based on a conceptual and theoretical approach to the safety of

large, complex, socio-technical systems developed by Reason (1997, 2000). This brief account of

ICAM is taken directly from that of Gibb et al. (2004). The specific objectives of investigations

carried out using ICAM are to:

• establish all the relevant and material facts surrounding the event,

• ensure that the investigation is not restricted to the errors and violations of operating

personnel,

• identify the underlying or latent causes of the event,

• review the adequacy of existing controls and procedures,

• recommend corrective actions to reduce risk, prevent recurrence and improve operational

efficiency,

• detect developing trends that can be analysed to identify specific or recurring problems,

• ensure that it is not the purpose of the investigation to apportion blame or liability, and

• meet relevant statutory requirements for incident investigation and reporting.

Reason (1997) defines organisational accidents as those in which latent conditions (arising

mainly from management decisions, practices or cultural influences) combine adversely with local

triggering conditions and with active failures (errors and/or procedural violations) committed by

individuals or teams to produce an accident. The Reason Model and ICAM focus on those matters

over which management could reasonably have been expected to exercise some control.

The ICAM Model organises incident causal factors into four elements as illustrated in Figure

2. Organisational factors are the underlying factors which promote the task/environmental

conditions that affect performance in the workplace, or allow those conditions to remain

unaddressed. They may lie dormant or undetected for some time, and their repercussions may

become apparent only when they combine with local conditions and errors to breach the system’s

defences. Task/environmental conditions are the task, situational or environmental conditions in

existence immediately before, or at the time of, the incident. They are the circumstances under


which the errors or violations took place. They can be embedded in the demands of the task, the

work environment, individual capabilities and/or human factors. Individual or team actions are

errors or violations of a standard or procedure committed in the presence of a potential hazard that

is not properly controlled. They are the actions or omissions that led directly to the incident or

accident. Absent or failed defences fail to provide the required protection to the system against

technical or human failures. Both proactive and reactive defences are required to prevent incidents

or accidents and to minimise adverse effects if they do occur (Gibb et al., 2004).

Fig. 2. The ICAM model of accident causation (Gibb et al., 2004).

4.3. System Safety Accident Investigation (SSAI)

The System Safety Accident Investigation (SSAI) approach was developed in the 1970s for use in

the US nuclear power industry. The techniques have since been modified for use in many high risk

industries where fatalities or other catastrophes can occur. This brief account of SSAI is drawn

from that given by Joy (2004) who has developed the approach particularly for application to the

mining industry.

SSAI provides a systematic and logical process for fact finding in accident investigations and

the drawing of conclusions. It imposes an overall discipline on the investigation process, providing

a systematic method of identifying what happened and why it happened. SSAI uses several

analytical techniques which are usually applied in a specific order:

• event and condition charting for displaying graphically the events in the accident sequence

and the preconditions that affected those events,

• fault tree analysis for depicting the possible scenarios leading to an event in the accident

sequence (where there were no witnesses),

Sound organisational

factors

Produce favourable conditions

Reduce errors and violations

Safety net Redundancy

Risk management Error traps

Error mitigation

Safe and

successful task

completion

Organisational factors

Task / environment conditions

Individual / team

actions

Absent / failed defences

Staff selection

Training Procedures

Equipment selection Equipment design

Operations vs safety goals

Contractor management Management of change

Working

conditions Time pressures

Resources Tool availability

Job access Task complexity Fitness for work

Workload Task planning

Errors and

violations

Interlocks Isolation Guards Barriers

Safe operating procedures Job safety analyses

Awareness Supervision

Emergency response Personal protective

equipment

Accident

Incident

Near miss

Equipment failure


• energy/barrier analysis to illustrate the unwanted energy flows and barrier inadequacies that

contributed to the accident,

• human error analysis for the systematic examination of any deviations from expected human

performance, and

• gap analysis to provide some insight into why the accident occurred by comparing accident-

free conditions with the accident conditions.

Not all of these techniques are necessarily applied in any given accident investigation. They

are tools to be used by an accident investigation facilitator. Typically, accidents resulting in or

having the potential for loss of life, massive equipment damage, or prolonged system failure,

receive a full and extensive investigation. The author has been part of such SSAI investigations in

two cases involving fatalities in the Australian mining industry.

5. The Legal Interface

It is axiomatic to the concept and definition of forensic engineering being used here, that the

forensic engineer’s investigations have a relation to the legal system of the state or country in

which the incident took place. The forensic engineer may be required to act as an expert witness in

coronial enquiries, in civil or criminal court proceedings, in mediated or arbitrated disputes, or in

tribunals. Even in legal cases that may be settled out of court, the forensic engineer will usually be

required to prepare an expert witness statement in a prescribed format.

Most court systems issue guidelines for the preparation of expert witness statements,

emphasising the need for the expert witness to be objective and impartial in giving opinion

evidence (e.g. Department of Constitutional Affairs, UK, 2006; Federal Court of Australia, 2004).

In the adversarial court systems in which the author has appeared as an expert witness, cross-

examination by barristers can be a challenging, and sometimes harrowing, experience. The Federal

Court of Australia’s guidelines make provision for the Court to direct the experts retained by the

parties to meet in an attempt to reach agreement about matters of expert opinion. Interestingly, the

Civil Procedure Rules of the UK now give the Court the power to direct that evidence on a given

issue be given by a single joint expert (Department of Constitutional Affairs, UK, 2006).

Professional engineering practice is governed by Codes of Ethics to which engineers subscribe

on becoming members of professional engineering organisations. Ethical conduct is especially

important in forensic engineering where the stakes can be high and there can be pressure on the

forensic engineer to orient his/her investigation, findings and/or expert statement in the interests of

one party or another. The Guidelines for Forensic Engineering Practice of the American Society

of Civil Engineers (Lewis, 2003) define ethical forensic engineering practice as “the conduct of

forensic investigations and providing expert testimony based on sound, comprehensive, and

unbiased investigation, and demonstrating exemplary professional conduct and honesty in serving

the trier of fact, the public, and clients, as a qualified expert”.

A particular area of litigation often arising in underground construction concerns the

consequences of what may be described variously as unforeseen, changed, differing or latent

ground conditions. These are ground conditions encountered during construction that it is argued

could not reasonably have been foreseen at the time of tendering. Modern risk management

techniques and contractual practices (e.g. Eskesen et al., 2004) seek to avoid or minimise litigation

arising from this cause. However, the area remains one in which forensic geotechnical engineers

often become seriously involved. Gould (1995) notes that, in the USA, progress in dispute

resolution has brought with it the increasing involvement of geotechnical engineers in changed-


condition claims. Gould (1995) further notes that “in the present highly competitive heavy-

construction market, bidders commonly accept hazards without contingency in their proposal,

expecting that “claimsmanship” will turn a marginal job into a profitable one”.

In some jurisdictions, proceedings in underground construction and mining cases may be

taken by government agencies under occupational or work place health and safety legislation. In

some states of Australia there appears to be an increasing trend to institute such proceedings

against individual professional engineers and against operating companies, contractors and

suppliers, for exposing employees to risk in the work place, both in the civil construction and

mining industries. Despite the significant advances that have been made in the last decade or more

in reducing geotechnically-related lost time injury and fatality rates in Australia’s underground

mines, Galvin (2005, 2006) has found that seemingly automatic prosecution policies (and some

court decisions) are now impacting negatively on the objective of reaching the goal of zero harm

because:

• “Lessons from serious incidents are not being disseminated until some years after because of

privilege and other considerations associated with pending charges.

• Some organizations and employers are reluctant to encourage near-miss reporting because

of concerns that it could be used against them in prosecutions.

• It is a major disincentive for young people to seek a management career in the minerals

industry.”

6. The Lane Cove Tunnel Project

6.1. Background

The Lane Cove Tunnel Project (LCTP) in Sydney, New South Wales, Australia, involves the

construction of twin, 3.6 km long, two- and three-lane tunnels together with 3.5 km of bridge and

road upgrades to link the M2 motorway at North Ryde with the Gore Hill Freeway, as well as a

number of other elements. The twin east-bound and west-bound tunnels run beneath and slightly to

the north of Epping Road. Figure 3 shows a schematic diagram of the tunnels in the project.

Fig. 3. Tunnel schematic, Lane Cove Tunnel Project (Rozek, 2005).

Mowbray Park work site

Sirius Road Work site

Pacific Highway exit ramp

Eastbound tunnel

Westbound tunnel

Pacific Highway entry ramp

Pacific Highway entry ramp

Fresh air intake

Marden Street work site

Mid-tunnel access

Collapse at intersection

Eastbound tunnel

Westbound tunnel


The New South Wales Roads and Traffic Authority (RTA) engaged the Lane Cove Tunnel

Company (LCTC) to design, construct, maintain and operate the tunnel for a period of 33 years.

LCTC, in turn, appointed the Thiess John Holland Joint Venture (TJH) to design and construct the

project. TJH’s contract commenced in December 2003 and tunnel construction in July 2004. The

tunnel works are divided into three parts – a western or Mowbray Park section, a central section,

and an eastern or Marden Street section. The contract completion date is May 2007.

The majority of the LCTP tunnelling is in the Hawkesbury Sandstone whose properties and

engineering behaviour have been the subject of detailed study. Those sections of the tunnelling to

be discussed here are excavated in the overlying and generally horizontally-bedded Ashfield Shale,

described by Badelow et al. (2005) as a sequence of “mudrocks including siltstone, mudstone or

laminate and lesser shale and claystone”. In the process of the investigation, design and

construction of a range of excavations and foundations in the Triassic rocks of the Sydney region,

local classifications of the sandstones and shales have been developed and applied (e.g. Bertuzzi

and Pells, 2002b; Pells et al., 1998). Specific design methods for the excavations in the Sydney

rocks and for their support and reinforcement have also been developed and applied with notable

success (e.g. Bertuzzi, 2005; Bertuzzi and Pells, 2002a; Pells, 2002; Pells et al., 1991). Advantage

was taken of these methods and this experience in the design of the LCTP tunnels (Badelow et al.,

2005; Maconochie et al., 2005; Rosek, 2005).

In the early hours of Wednesday, 2 November 2005, subsidence developed above the Pacific

Highway Exit Ramp tunnel on design control line MCAA at its junction with the Marden Street

ventilation tunnel on design control line MC5B. The road-header and loader working at the

location were buried by collapsed material. The subsidence propagated to surface near 11-13

Longueville Road, Lane Cove, and an exit ramp from Longueville Road at the point marked in

Figure 3. The subsidence under-mined the front of a building (producing a spectacular effect) and

a bored pile retaining wall on the north side of Longueville Road shown in Figure 4. Within a few

hours of the collapse, it was decided to fill the cavity with concrete as illustrated in Figure 4 in

order to arrest the under-mining and underpin the retaining wall. The incident attracted

considerable media attention in Sydney and more widely in Australia.

6.2. The Investigation

Two days after the incident, the author was engaged, through Golder Associates Pty Ltd, as an

independent expert to investigate and report on the cause(s) of the subsidence. The author visited

Sydney in the period 7-11 November 2006 when he carried out the following main activities:

• discussions and interviews with representatives of the contractor, the designer, the geotechnical

consultant, and the crew and supervisors working at the collapse site at the time of the incident,

• a surface site visit,

• an underground visit to the site of the collapse and to inspect other tunnels in the Marden Street

section of the project,

• study of a wide range of design documents, drawings, construction reports and other

documents, and

• giving preliminary consideration to the possible causes of the incident and the preparation of

an outline of the report.


Fig. 4. Vertical section through the Lane Cove Tunnel incident site following the concrete pours.

Following this preliminary analysis of the available information, the following possible

contributory factors to the collapse were identified:

• water,

• the weak shale in which the intersection was excavated,

• a low strength, steeply-dipping dolerite dyke crossing the intersection,

• jointing associated with the dyke,

• faulting present at the site,

• the large effective span of the intersection excavation,

• the low depth and nature of the cover,

• the levels of support provided, and

• the proximity of the Longueville Road retaining wall and ground anchors to the roof of the

intersection excavation (see Figure 2).

The author then returned to Brisbane to continue his analysis of the information collected

during the initial site visit and to prepare his report. During this period, several further documents

and items of information were requested and supplied by the contractor and the geotechnical

consultant. In the analysis of the incident, formal investigation techniques of the types outlined in

Section 4 above were not utilised. However, the issues considered in the SSAI approach were

addressed in the investigation, although not in a formal manner. A sequence of events was

compiled in tabular rather than diagrammatic format.

The final report (Golder Associates, 2005) was released to the public and attracted some

attention in the Sydney media. Other reports of the incident, generally based on the author’s report,

appeared in the technical press (e.g. Kitching, 2006). A few months later, the WorkCover

Authority of New South Wales made public an initial report of its own investigation of the incident

17 m

25 m

5 MPa flowable concrete

50 MPa concrete

Unconsolidated fill

DatuDatum R.L. 69.849

4.54 m 6.00 m 15.72 m 11.88 m

Intersection of debris with tunnel roof

Ventilation tunnel Marden Street MC5B

Existing surface

13 Longueville Road

Piles Anchor

Approximate line of land slip

3 m

4 m R.L. 94.0 Road

49.0

0 m

MC

1A

MC

2A

16.50 m 15.41 m

MC

AA

Datum R.L. 45.0


carried out pursuant to Section 88 of the NSW Occupational Health and Safety Act 2000

(WorkCover NSW, 2006).

6.3. Conclusions Drawn

The report into the causes of the subsidence (Golder Associates, 2005) drew the following

conclusions:

(i) Tunnelling and underground construction are always attended by a number of risks and

uncertainties, mainly associated with the inherent variability of the geological structure and

mechanical properties of the rock masses in which construction takes place.

(ii) During the excavation of the Marden Street ventilation tunnel, a near-vertical dolerite dyke

(or pair of dykes), was intersected at a number of locations. Although dykes are known to

exist in the sedimentary rocks of the Sydney region, this particular dyke had not been

identified in the pre-construction geotechnical investigation. However, the dyke had been

intersected previously during construction in ventilation tunnel MC5A and the main lane

tunnels MC1A and MC2A.

(iii) Because of the presence of two sets of orthogonal joints associated with the dyke and other

jointing and faulting, the shale rock mass at and near the junction of the MC5B ventilation

tunnel and the MCAA exit ramp was of poorer quality than had been anticipated in the

design stage.

(iv) The collapse started near the north-west corner of the newly extended down drive of the

MCAA exit ramp at about 1:38 am on the morning of 2 November 2005, and rapidly

extended across the exit ramp face to the dyke in the crown. The fall propagated across the

crown of the junction of the MCAA exit ramp and the MC5B ventilation tunnel to the east or

south-east and included the dyke. The collapse propagated to surface in Longueville Road in

10-20 minutes.

(v) The collapse initiated with the fall of rock blocks in the north-west corner of the excavation

as a result of unravelling under a lack of lateral and normal restraint.

(vi) The ultimate failure mechanism was progressive, probably consisting of several stages (listed

in the report).

(vii) The processes and methodology used in the design of the LCTP tunnels was in accord with

best practice in Sydney and elsewhere, and the resulting designs were generally suitable for

their purposes.

(viii) In the design stage, no special analysis of the MC5B/MCAA junction was carried out.

However, because of the inevitable local variations in the geological and geotechnical

conditions, it was recognised that it would be necessary to modify or adapt the initial design,

particularly the support provisions, to the conditions actually encountered during

construction.

(ix) TJH has in place a series of appropriate and best practice processes for the safe and

productive execution of the underground construction works on the LCTP. Some of the

documents setting out these processes are models of their kind.

(x) Up to the time of the incident of 2 November 2005, the designs and processes in place had

been executed in a highly professional and productive manner by a knowledgeable and

dedicated workforce and their supervisors.


(xi) The collapse arose from a combination of factors that were not present together at any other

location in the underground works on the project.

(xii) The factors causing the collapse were probably:

• the presence of the dyke providing a persistent, relatively low strength, near vertical

discontinuity transecting the roof of the excavation in a strike direction that was closely

parallel to the maximum effective span of the junction,

• the presence of orthogonal, closely-spaced jointing associated with the dyke, reducing

the already poor mechanical quality of the weathered Ashfield Shale rock mass,

• the presence of faults with orientations such that, in conjunction with the dyke, the joints

and the excavations boundaries, they could isolate blocks that were free to fall or slide

from the excavation boundaries if not adequately supported,

• a large effective span with relatively low cover to rock head, and

• the level of support existing in the western side of the excavation at the time being

inadequate to ensure the excavation’s stability given the large effective span, the low

rock cover, the presence of persistent vertical discontinuity (the dyke) transecting the

excavation, and the poor mechanical properties of the overlying rock mass.

(xiii) Water was not a cause of the collapse.

(xiv) The proximity of the Longueville Road retaining wall and its ground anchors to the crown of

the MC5B/MCAA junction excavation may have contributed to the collapse by influencing

the loads applied to the rock immediately above the excavation, or by weakening that rock

mass, or both.

(xv) The preparation of the best possible longitudinal geological sections and/or progressive

geological plans may have been of assistance in projecting conditions ahead of the face as

excavation progressed.

7. Concluding Remarks

Forensic investigations of underground construction and mining failures and the associated

interactions with the legal system, now represent an important part of the work of many

experienced geotechnical engineers. The conduct of these investigations, the drawing of

conclusions and the communication of the results impose significant demands on the knowledge

and skills of the forensic geotechnical engineer and require the exercise of the highest professional

and ethical standards. Forensic geotechnical engineers can play a significant role by explaining

publicly and within the legal system, the not widely understood difficulties and risks involved in

engineering in the variable natural materials found on and in the crust of the Earth.

Acknowledgments

The author wishes to thank the management and staff of the Brisbane office of Golder Associates

Pty. Ltd, for their support and assistance in the preparation of this paper. The author also wishes to

thank his former colleague at the University of Queensland, Professor Jim Joy, for introducing him

to the SSAI approach and for providing him with information used in the preparation of this paper.

Finally, the permission given by the Thiess John Holland Joint Venture to publish details of the

author’s investigation of the Lane Cove Tunnel Project incident is gratefully acknowledged.


References

Alonso, E. E. and Gens, A. (2006a). “Aznalcóllar Dam Failure. Part 1: Field Observations and

Material Properties”. Géotechnique, 56(3): 165-183.

Alonso, E. E. and Gens, A. (2006b). “Aznalcóllar Dam Failure. Part 3: Dynamics of the Motion”.

Géotechnique, 56(3): 203-210.

Badelow, F., Best, R., Bertuzzi, R. and Maconochie, D. (2005). “Modelling of Defect and Rock

Bolt Behaviour in Geotechnical Numerical Analysis for Lane Cove Tunnel”. Proceedings of

the Mini-Symposium: Geotechnical Aspects of Tunnelling for Infrastructure Projects, Sydney

Chapter, Australian Geomechanics Society, 12 October, 2005.

Bea, R. (2006). “Reliability and Human Factors in Geotechnical Engineering”. Journal of

Geotechnical and Geoenvironmental Engineering, ASCE, 132(5): 631-643.

Bertuzzi, R. (2005). “Tunnelling in Sydney: the Next Geotechnical Step”. Proceedings of the

Mini-Symposium: Geotechnical Aspects of Tunnelling for Infrastructure Projects, Sydney

Chapter, Australian Geomechanics Society, 12 October, 2005.

Bertuzzi, R. and Pells, P. J. N. (2002a). “Design of Rock Bolt and Shotcrete Support of Tunnel

Roofs in Sydney Sandstone”. Australian Geomechanics, 37(3): 81-90.

Bertuzzi, R. and Pells, P. J. N. (2002b). “Geotechnical Parameters of Sydney Sandstone and

Shale”. Australian Geomechanics, 37(5); 41-54.

Brookes, C. (2006). “Forensic Engineers – Do We Want to be Heard?” Materials World, 14(4):

34-36.

Carper, K. L. (ed.) (2001). Forensic Engineering, 2nd

edition. CRC Press LLC: Boca Raton,

Florida.

Christian, J. T. (2004). “Geotechnical Engineering Reliability: How Well do We Know What We

are Doing?” Journal of Geotechnical and Geoenvironmental Engineering, ASCE, 130(10):

985-1003.

Day, R. W. (1999). Forensic Geotechnical and Foundation Engineering. McGraw-Hill: New

York.

De Ambrosis, L. P. and Kotze, G. P. (2004). “Stress Induced Roof Collapses During Construction

of the Sydney LPG Storage Cavern”. Proceedings of the 9th

Australia New Zealand

Conference on Geomechanics, Auckland, 8-11 February, 2004, 1: 159-165.

Department of Constitutional Affairs, UK (2006). Civil Procedure Rules Part 35: Experts and

Assessors. Available at http://www.dca.gov.uk/civil/procrules_fin/contents/parts/part35.htm .

Einstein, H. H. (1996). “Risk and Risk Analysis in Rock Engineering”. Tunnelling &

Underground Space Technology, 11(2): 141-152.

Eskesen, S. D., Tenborg, P., Kampmann, J. and Holst Veicherts, T. (2004). “Guidelines for

Tunnelling Risk Management: International Tunnelling Association, Working Group No 2”.

Tunnelling and Underground Space Technology, 19(3): 217-237.

Federal Court of Australia (2004). Practice Direction: Guidelines for Expert Witnesses in

Proceedings in the Federal Court of Australia, March 2004. Available at

http://www.fedcourt.gov.au/how/prac_direction.html .

Feld, J. and Carper. K. L. (1997). Construction Failure, 2nd

edition. John Wiley & Sons: New

York.

Galvin, J. M. (2005). “Occupational Health and Safety Acts – Performance and Prosecution in the

Australian Minerals Industry”. Mining Technology (Transactions of the Institution of Mining

& Metallurgy, Section A), 114: A251-A256.


Galvin, J. (2006). “Health and Safety in Australia’s Mines”. Materials World, 14(4): 22-23.

Gens, A. and Alonso, E. E. (2006). “Aznalcóllar Dam Failure. Part 2: Stability Conditions and

Failure Mechanism”. Géotechnique, 56(3): 185-201.

Gibb, G., Reason, J., De Landre, J. and Placanica, J. (2004). “The Incident Cause Analysis

Method (ICAM)”. Safety in Australia, 26(2): 13-19.

Golder Associates (2005). Causes of Subsidence, 2 November 2005, Lane Cove Tunnel Project,

Sydney, NSW. Report No 001-05632178 to Hon. Andrew Rogers QC, Sydney, by Golder

Associates Pty Ltd, Brisbane.

Gould, J. P. (1995). “Geotechnology in Dispute Resolution”. Journal of Geotechnical

Engineering, ASCE, 121(7): 523-534.

Greenspan, H. F., O’Kon, J. A., Beasley, K. J. and Ward, J. S. (1989). Guidelines for Failure

Investigation. ASCE: New York.

Hadipriono, F. C., (2002). “Forensic Study for Causes of Fall Using Fault Tree Analysis”. Journal

of Performance of Constructed Facilities, ASCE, 15(3): 96-103.

Hoek, E. (2001). “Big Tunnels in Bad Rock”. Journal of Geotechnical and Geoenvironmental

Engineering, ASCE, 127(9): 726-740.

Hudson, J. A. (2006). Personal communication.

Joy, J. (2004). “Occupational Safety Risk Management in Australian Mining”. Occupational

Medicine, 54(5): 311-315.

Kitching, R. (2006). “Short Rock Bolts Blamed for Tunnel Cave-in”. New Civil Engineer

International, March: 4-5.

Kontogiannis, T., Leopoulos, V. and Marmaras, N. (2000). “A Comparison of Accident Analysis

Techniques for Safety-Critical Man-Machine Systems”. International Journal of Industrial

Ergonomics, 25: 327-347.

Leonards, G. A. (1982). “Investigation of Failures”. Journal of the Geotechnical Engineering

Division, ASCE, 108(GT2): 187-246.

Lewis, G. L. (ed.) (2003). Guidelines for Forensic Engineering Practice. ASCE: Reston, Virginia.

Lewis, P. R. (2004). Beautiful Railway Bridge of the Silvery Tay. Tempus: Stroud, U. K.

Londe, P. (1987). “The Malpasset Dam Failure”. Engineering Geology, 24(1-4): 295-329.

Maconochie, D. J., Loganathan, N. and Bertuzzi, R. (2005). “Design of the Lane Cove Tunnel,

Sydney”. Proceedings of the 12th

Australian Tunnelling Conference, Brisbane, 17-20 April,

2005, 587-603.

Müller, L. (1968). “New Considerations of the Vajont Slide”. Rock Mechanics and Rock

Engineering, 6(1): 1-91.

Noon, R. (2001). Forensic Engineering Investigation. CRC Press LLC: Boca Raton, Florida.

Osterberg, J. O. (1989). “Necessary Redundancy in Geotechnical Engineering”. Journal of

Geotechnical Engineering, ASCE, 115(11): 1513-1531.

Pells, P. J. N. (2002). “Developments in the Design of Tunnels and Caverns in the Triassic Rocks

of Sydney Region”. International Journal of Rock Mechanics and Mining Sciences, 39(5):

569-587.

Pells, P. J. N., Mostyn, G. and Walker, B. F. (1998). “Foundations on Sandstone and Shale in the

Sydney Region”. Australian Geomechanics. 33(3): 17-29.

Pells, P. J. N., Poulos, H. G. and Best, R. J. (1991). “Rock Reinforcement and Design for Shallow

Large Span Cavern”. Proceedings of the 7th

Congress, International Society for Rock

Mechanics, Aachen, September 16-20, 1991, 2: 1193-1198.


Petroski, H. (1985). To Engineer is Human: The Role of Failure in Successful Design. Macmillan:

London.

Petroski, H. (1994). Design Paradigms: Case Histories of Error and Judgement in Engineering.

Cambridge University Press: Cambridge.

Potts, D. M., Dounias, G. T. and Vaughan, P. R. (1990). “Finite Element Analysis of Progressive

Failure of Carsington Embankment”. Géotechnique, 40(1): 79-101.

Reason, J. (1997). Managing Risks of Organisational Accidents. Ashgate: Aldershot, U. K.

Reason, J. (2000). “Human Error: Models and Management”. British Medical Journal, 320(7237):

768-770.

Rosek, J. (2005). “Delivery of the Epping Road to Chatswood Rail Line and Lane Cove Tunnels”.

Proceedings of the 12th

Australian Tunnelling Conference, Brisbane, 17-20 April, 2005.

Shirlaw, J. N. and Boone, S. (2005). “The Risk of Very Large Settlements Due to EPB

Tunnelling”. Proceedings of the 12th

Australian Tunnelling Conference, Brisbane, 17-20

April, 2005.

Shirlaw, J. N., Ong, J. C. W., Rosser, H. B., Tan, C. G., Osborne, N. H. and Heslop, P. E. (2003).

“Local Settlements and Sinkholes due to EPB Tunnelling”. Proceedings of the Institution of

Civil Engineers, Geotechnical Engineering, 156(GT4): 193-211.

Skempton, A. W. and Vaughan, P. R. (1993). “The Failure of Carsington Dam”. Géotechnique,

43(1): 151-173.

Sowers, G. F. (1993). “Human Factors in Civil and Geotechnical Engineering Failures”. Journal of

Geotechnical Engineering, ASCE, 119(2): 238-256.

Specter, M. M. (1987). “National Academy of Forensic Engineers”. Journal of Performance of

Constructed Facilities, ASCE, 1(3): 145-149.

Stabel, B. and Samani, F. B., 2003. “Masjed-e-Soleiman HEPP, Iran: Rock Engineering

Investigations, Analysis, Design and Construction”. Proceedings of the 10th

Congress,

International Society for Rock Mechanics, Johannesburg, 8-12 September, 2003, 2: 1147-

1154.

Terzaghi, K. (1929). “Effect of Minor Geological Details on the Safety of Dams”. American

Institute of Mining and Metallurgical Engineers, Technical Publication 215, 31-44.

Wearne, P. (1999). Collapse: Why Buildings Fall Down. Channel 4 Books: London.

WorkCover NSW (2006). Lane Cove Tunnel Collapse and Subsidence, 2 November 2005. Report

of the WorkCover Authority of New South Wales, 27 March 2006.

Documents

Forensic Engineering Underground